Venkatesh

Out in Imaging Neuroscience!
direct.mit.edu/imag/article...

Out in Imaging Neuroscience!
direct.mit.edu/imag/article...

The broken link is supposed to be youtu.be/VCH34X6l2ao

The animations above are made possible thanks to the superb Manim community, massive shoutout to them, and the legend Grant Sanderson !
Fin.

There were really a lot of little manoeuvres costing us 51 years rofl, but thrilled AF this work sees the world !!!! teamwork with my supervisor gang Giulia Lioi, Karim Jerbi, and @nicolasfarrugia.bsky.social

Together, the characterization between structure and function highlights the role of anatomy in shaping ongoing cognitive processes, not only in specific brain areas but also in task conditions.
More details in the preprint!! www.biorxiv.org/content/10.1...

We also tested if the coupling relationship is specific to frequency bands (cross-f analysis) and found similar patterns across EEG frequency bands. Specifically, the macro-scale relationship (Yeo-Krienen networks) was highly similar.

How about the temporal axis? To test the within-condition differences, we selected moments of High (and low) Inter-subject correlation, extending the work from Uri Hasson among many others. We found that salient events in the movie do not modulate the coupling relationship.

We wondered if the task (video) engagement modulates the cortical EEG's relationship with anatomy. Turns out yes; we found differences during video and rest, with part of the sensorimotor cortex exhibiting strengthened coupling to anatomy.

We found that the wideband EEG in the sensorimotor cortex is strongly constrained by the anatomy, with higher-order systems exhibiting flexibility, a trend consistent across multiple videos and resting-state.

Results time! We observed brain-wide coupling and focal decoupling across videos and rest. We performed a meta-analysis (functional decoding) with the existing literature using NiMare/Neurosynth to interpret the results.

In a purely GSP lingo, we put the signal onto the graph and assess how smooth the signal is (vs "roughness"). When the graph signal is source-localized EEG and the graph is anatomical connectivity, this tells us how the EEG propagates thru the anatomical connections.

Eigenmodes here are atoms/notes/alphabets of the anatomy that any activity (element/chords/words) can be expressed with by a weighted linear combination. Thus, brain activity can be turned into structurally informed components, i.e activity as an extension of the anatomy!->https://youtu.be/VCH34X6l2

Plucking a violin string induces oscillations (aka harmonics). Decomposing the structural connectome can be thought of the same way describing harmonics in the connectome--vibrations in the spatial axis. -> youtu.be/yb8UZmnHx5I
Credits: Müller et al '22 for the💡

GSP is an elegant framework that's been used here to decompose the structural graph into eigenmodes and express the activity in a structurally informed way. Here are some smooth animations to help you visualize its ubiquitous nature and how they're adapted to study brain activity

Extending the previous works such as Glomb et al 2020, Rué-Queralt et al. 2021, Griffa and Preti 2022 and by applying the GSP metric developed by
Preti and Van de Ville (2019), we assessed the region-level association between continuous EEG and anatomy

Traditionally often, the activity and the connectivity are studied in a vacuum. Here we take advantage of two complementary domains and study them jointly.
We are interested in how the activity propagates on the structural pathways, aka Structure-Function coupling. youtu.be/yKb5Ue2f7rI

The end of 3 yrs of "ah this looks reassuring, but,"
biorxiv.org/content/10.1... let's dive in🧵
We investigated the link b/w source-localized continuous EEG (Movie &Rest) and individual anatomical scaffold using Eigenmodes / Graph Signal Processing(GSP) #neuroskyence

#HiSciSky I am a PhD candidate between Université de Montréal and IMT Atlantique France, aiming to characterising the interplay between structure and function across various conscious states. Eigenmodes / Graph Signal Processing #neuroscience #consciousness
Looking forward to excited science here !